The Effectiveness of an Augmented Reality (AR) Training Module: A Kirkpatrick-Based Framework from Development to Deployment

Abstract

Augmented Reality (AR) in education has become popular as a novel resource to boost engagement and facilitate experiential learning. Nonetheless, numerous educators have limited experience with AR development tools, which hinders their ability to integrate this technology into their teaching effectively. This study aimed to evaluate the effectiveness of a step-by-step AR Training Module, designed using Unity and Vuforia, through Kirkpatrick’s Four-Level Model: Reaction, Learning, Behaviour, and Results. A structured questionnaire of 20 Likert-scale items was administered to 37 lecturers from Politeknik Muadzam Shah and selected Kolej Komuniti institutions. The reliability test was assessed using Cronbach’s Alpha, which yielded values between 0.883 and 0.928, indicating high reliability. Descriptive statistics were calculated to determine each Kirkpatrick level's mean and standard deviation, revealing strong agreement on the module’s effectiveness, particularly in the Reaction (M = 4.43, SD = 0.54) and Results (M = 4.08, SD = 0.71) domains. The Learning level also showed positive outcomes (M = 3.94, SD = 0.57), while the Behaviour level indicated moderate agreement (M= 3.70, SD = 0.63) regarding the application of acquired skills. These findings support the AR module's ability to deliver engaging, relevant content with measurable impact across different dimensions of learning effectiveness. Independent sample t-tests were performed to evaluate perceptions between institutions, revealing no statistically significant differences, indicating that the module is similarly practical across various educational environments. All statistical evaluations were performed utilising IBM SPSS Statistics version 31.0. The findings demonstrate that the AR module successfully assists beginner users in gaining essential skills in AR development and is positively regarded in diverse institutional settings. The lack of notable differences between institutions indicates that the module is flexible and can be expanded for broader application. This research contributes to the expanding literature on AR integration within education and presents a validated framework for assessing comparable digital training initiatives moving forward.